Mobile radio communications suffer from large-scale and small-scale fading effects that attenuate the communication signal. While large-scale fading is caused by distance-dependent path loss and shadowing effects, small-scale fading is caused by multipath propagation. For mobile receivers or transmitters, small-scale fading can cause rapid fluctuations of the received signal-to-noise ratio (SNR); if a mobile device moves only a small distance it may experience deep fading even if it had perfect signal reception just an instant before.
Cooperative relaying [1] is a concept, where a relay node assists the communication between two nodes when the direct link is affected by fading. The information is relayed via a spatially different path which is likely not affected by the same fading effects as the direct link at the same time. Thus, using such a relay communication channel can improve the communication performance by implementing spatial diversity for the communication paths [12]. With the growing number of networked wireless devices in everyday appliances, there are more potential relay nodes within transmission range of a sender and receiver. Henceforth, cooperative relaying will gain additional importance in the near future.
Cooperative diversity is expected to be more beneficial, if the cooperative relaying protocol is designed according to the following: First, it should have a low overhead, for example in terms of sent messages, energy, storage, processing. A large number of communication attempts are expected to succeed without the need for alternative communication paths. Thus, in the case of a successful transmission, a cooperative relaying protocol should have minimal overhead in comparison to non-cooperative transmission schemes. Second, the protocol should exploit cooperative diversity to an extent that makes the effort for the more complex interaction between wireless nodes worth it.
Cooperative relaying can be divided into three main phases: direct transmission, relay selection, and cooperative transmission. In the direct transmission phase the source transmits its data, whereas destination and relay (or potential relays) try to receive it. In the relay selection phase a neighboring node of source and destination is selected. The cooperative transmission phase, where the relay forwards the data to the destination, occurs only if the destination has failed to retrieve the data from the source during the direct transmission.
The relay selection phase has a great impact on the performance of the whole cooperative relaying process [5]. The major selection criterion is the link quality of the communication participants which is typically measured by probing packets [6, 7, 10]. The selection can be further refined by using additional factors like residual power [8]. Note that the selection process also depends on the actual environment, i.e., it is important to know how frequently a relay needs to be selected for a given source destination pair, since a node may be a good relay at one time instant and a bad one later on. In the absence of any environment information, relays are selected for each packet anew [10]. Typically, relay selection is a distributed task which necessitates time and energy and thus introduces additional overhead. Therefore, it is beneficial to explore the current realization of the channel between source and destination and to do relay selection only on demand [9].
Relay selection can be done before direct transmission (proactive relay selection) or after direct transmission (reactive relay selection). Proactive relay selection is considered to have energy advantages over reactive relay selection since only the selected relay needs to spend energy for listening to the transmission of the source [10]. However, it introduces a constant overhead to all transmissions. Moreover, the channel state may change during the direct transmission phase. The selected relay might then not be able to receive the data, or the relay link to the destination can also change considerably making successful relaying unlikely. In reactive schemes, a relay is selected only if the destination is not able to receive the data from the source and asks for assistance (cf. iterative relaying [12]). Thus, reactive relay selection is only done if a direct transmission fails and relaying candidates have already received the original data from the source properly. A disadvantage is that all potential relaying nodes need to listen to the transmission of the source. Since many transceivers consume the same order of energy for receiving and transmitting [11], the costs for having all neighbors of source and destination listen might not be negligible.
Recently, cooperative relaying is no longer treated as a separate task but is investigated in combination with MAC protocols. It is beneficial to exploit channel reservation messages such as Request-To-Send (RTS) and Clear-To-Send (CTS) for probing the channel and selecting a relay [13, 14], since relays also need to make a channel reservation for the cooperative transmission phase. The herein proposed approach selects relays after the direct data transmission phase and thus, does not introduce additional delay to successful direct transmissions. In [13] neighbors determine whether cooperation is necessitated based on the channel quality between source and destination. Relay selection and the relay's channel reservation are done by using a slotted approach before the direct transmission.
Liu et al. introduce in [14] a protocol which uses information from the past, to determine whether the data transfer rate between two nodes can be improved by using a relaying node. Channel reservation for source, destination, and relay are done before direct transmission.
Moh et al. uses in [15] pro-active relay selection based on historical data and thus favor nearby neighbors as relay. Their protocol extends CSMA/CA to use distributed standard block codes (D-STBC) when needed, which allows the simultaneously transmission from two or more nodes. The cooperation is not just used for data transfer but also for signalling messages like RTS and CTS.
In summary, disadvantages of known relaying protocols are low flexibility, waste of resources including transmission resources/bandwidth on the one hand and battery power of wireless devices on the other hand due to the overhearing/storing of messages in relay devices, which would not have been necessitated.